December 5, 1SS4. 



SCIENCE. 



509 



est and most permanent usefulness, only when 

 it fulfils this intention as far as possible. 



Whether the words we have quoted, and 

 others of a similar tenor, mark a change of 

 opinion on the part of the director of the New- 

 York station, or are only a clearer expression 

 of convictions previously held, we do not un- 

 dertake to say. In either case, we are glad 

 to see the weight of this important institution 

 cast in favor of the scientific conception of an 

 experiment-station. The great need of agri- 

 culture to-da}' is not new varieties of plants, 

 or improved breeds of animals ; new methods 

 of cultivating the soil, or improved systems of 

 farming. All these, and many other like 

 things, are good ; but the two great wants 

 are a better knowledge of principles, and great- 

 er intelligence to applj'.them. For the latter 

 we must look to our agricultural schools : the 

 former we should require from our experiment- 

 stations. 



the laws and principles underlying agriculture ; 

 and, second, the experimental farm, devoted 

 mainly to carrying out upon a farming scale 

 the principles worked out by the experiment- 

 station. 



^Ye do not hold that an experiment-station 

 should never undertake to originate or test 

 new varieties of plants and animals or new 

 agricultural methods, — often work of this gen- 

 eral character will be demanded of it by the 

 public, and will prove of great public utility, — 

 but, in our view, it should not be allowed to 

 be, or to appear to be, the chief end of the 

 station. The two kinds of work are both 

 important, but we question the advisability of 

 attempting to unite them in one institution 

 and under one management. Each requires 

 facilities and talents peculiar to itself; and it 

 seems doubtful, whether, as a rule, one insti- 

 tution will be able to provide good facilities for 

 both kinds of experimentation, and still more 

 doubtful whether it can find combined in one 

 person the diverse knowledge and training re- 

 quired for their successful prosecution. With 

 the growth of agricultural experimentation 

 there might profitably be, we suspect, in the 

 majority of cases, a subdivision of it into 

 two overlapping 3-et independent classes. We 

 should have, first, the experiment-station prop- 

 er, aiming chiefly at a further elucidation of 



LETTERS TO THE EDITOR. 



**# Correspondents are requested to be as brief as possible. The 

 writer's name is in ail cases required, as proof of good, faith. 



Psychical research. 



Your issue of Oct. 17 contained two articles winch 

 are of good omen for the future of ' psychical re- 

 search' in America. Of the first, the editorial arti- 

 cle, I need say little. It is cordially welcomed by my 

 colleagues and myself for its recognition of the far- 

 reaching importance of an enterprise in the further 

 development of which our society will, we hope, go 

 hand in hand with yours. With the second article, 

 on 'psychic force,' our agreement is less complete; 

 but we still find nothing to complain of in the general 

 attitude of the distinguished writer. He, too, rec- 

 ognizes the legitimacy of the inquiry, while clearly 

 apprehending its difficulties. He describes with en- 

 tire justice the two opposed classes between which 

 psychical research has to clear a path, — the party of 

 easy credulity, and the party of easy incredulity; 

 and he points out with no more than proper empha- 

 sis the rigorous caution which every forward step 

 demands. Fraud and superstition have naturally 

 seized on what science has so systematically neg- 

 lected; and those who now endeavor to take the 

 subject up from the scientific side must accept the 

 fact and its consequences. 



So far, then, we are wholly at one with Professor 

 Newcomb; but we cannot quite so readily follow 

 him in his criticisms of our own doings. He begins 

 by condemning one of our public appeals for infor- 

 mation ; but his strictures seem to assume that all the 

 information which the appeal brings in will be re- 

 garded by us as a safe basis for conclusions. The 

 appeal is, of course, merely a first step, for whicli it 

 would be difficult to imagine any effective substitute; 

 though I may mention that a very large amount of 

 our information comes to us through private chan- 

 nels. The sifting and treatment of the evidence 

 according to scientific canons must be a subsequent 

 labor, the rationale of which could not be set forth, 

 or even suggested, in the terras of a short advertise- 

 ment. And of this labor no portion is more impor- 

 tant than the one which we are glad to find Professor 

 Newcomb so explicitly recognizing. — the application 

 of the doctrine of chances. In all those branches of 

 our inquiry where questions of coincidence occur, it 

 is clearly essential to ascertain, as definitely as may 

 be, how far the coincidences may fairly be ascribed 

 to chance. We have taken, and are still taking, great 

 pains to obtain this definite information. Very wide 

 inquiries have been made; and the results, though 

 far from complete, may still, I think, claim decidedly 

 more validity, as a basis of computation, than Pro- 

 fessor Newcomb' s guess at what "any physician will 

 consider quite within the bounds of probability." 

 It would require more space than I can ask for, to 

 comment on Professor New-comb's numerical argu- 

 ment in detail. But I may remark that he seems to 

 confuse the argument by classing all together what 

 he calls 'dreams, illusions, visions,' etc.; at least, if 



